"Your emotions make you human. Even the unpleasant ones have a purpose. If you ignore them, they just get louder and angrier."-Sabaa Everyone can express face from their acts or demeanors. Through, those are made a emotion.

"Your emotions make you human. Even the unpleasant ones have a purpose. If you ignore them, they just get louder and angrier."-Sabaa Everyone can express face from their acts or demeanors. Through, those are made a emotion.

Macromolecules are large, complex molecules composed of covalently bonded monomer units, existing in different stereochemical configurations and topologies. As a result of such chemical diversity, representing, comparing, and learning over macromolecules emerge as critical challenges. To address this, we developed a macromolecule graph representation, with monomers and bonds as nodes and edges, respectively. We captured the inherent chemistry of the macromolecule by using molecular fingerprints for node and edge attributes. For the first time, we demonstrated computation of chemical similarity between 2 macromolecules of varying chemistry and topology, using exact graph edit distances and graph kernels. We also trained graph neural networks for a variety of glycan classification tasks, achieving state-of-the-art results. Our work has two-fold implications - it provides a general framework for representation, comparison, and learning of macromolecules; and enables quantitative chemistry-informed decision-making and iterative design in the macromolecular chemical space. Macromolecules are ubiquitous and indispensable, from constituting what we are made up of to being present in almost everything we use. As biological macromolecules, they form the basis of life, serving as drivers of survival and growth functions. As synthetic macromolecules, humans have engineered the composition and topology to design structural components, sensors, shape-memory materials, drugs, encode messages, and much more (Lutz et al., 2016; Romio et al., 2020; Boydston et al., 2020; Thompson & Korley, 2020).

In machine learning when we build a model for classification tasks we do not build only a single model. We never rely on a single model since we have many different algorithms in machine learning that work differently on different datasets. We always have to build a model that best suits the respective data set so we try building different models and at last we choose the best performing model. For doing this comparison we cannot always rely on a metric like an accuracy score, the reason being for any imbalance data set the model will always predict the majority class. But it becomes important to check whether the positive class is predicted as the positive and negative class as negative by the model.